Protein neurotoxins are among the most powerful tools in the arsenal for investigation of membrane receptors and are now used as well for treatment of disease. This grant seeks, first, to advance our method of creation of de novo toxins by targeting potassium and proton channels that lack these useful ligands; second, to use de novo toxins to assess the potential of Kv1.5 and Kv2.1 channels as targets for treatment of two diseases of broad public impact- Atrial Fibrillation (AF) and Type 2 Diabetes (T2D), respectively. Success on these two fronts will allow wider use of these tools to produce toxins for other targets and motivate analogous work directed towards other types of membrane receptors important to human health and disease. To date, toxins have been derived primarily from limited sources: crude venoms, venom gland gene cloning, and laboratory synthesis of known toxins. Therefore, most receptors of interest lack a recognized, specific, high-affinity ligand. This state-of-affairs is easily understood: neither the purpose of toxins in the wild nor screens of venoms favor target specificity. Here, these problems are avoided by cloning toxins based on desired attributes using a new, high- throughput method we developed to generate a selective inhibitor of Kv1.3 channel pore. Specific Aim 1 seeks to improve the method by creating highly-diverse libraries on native scaffolds that bind to ion channel pores or voltage sensors, first using purified ion channel receptors, and then using channels on cell surfaces so the many receptors that do not retain their structure when purified can also be targeted. A recently identified channel for protons (Hv1) is used here to develop tools for voltage sensors (and to test directly the postulated role of this orphan receptor in male fertility). Specific Aim seeks de novo toxins directed to the Kv1.5 potassium channel. This orphan receptor is widely pursued as a potential therapeutic target for AF and lack of specific blockers has left unclear fundamental questions about its role in the heart and precluded direct tests of its utility as a drg target. Specific Aim 3 seeks de novo toxins specific for the Kv2.1 channel. Kv2.1 inhibition enhances glucose-stimulated insulin secretion from pancreatic ?-cells in a SUMO-dependent manner and inhibitors that do not cross inhibit are sought as potential drugs for T2D. De novo toxins for Kv1.5 and Kv2.1 are not only useful to validate the channels as drug targets but offer potential as therapeutic leads. Thus, small, disulfide-stabilized peptide toxins in use as drugs (Byetta(r), Integrilin(r), and Prialt(r)) have proven effective, protease resistant and non-immunogenic.